Abstract After a discussion of former publications on the problem of the existence of promethium in nature, studies o f our own on this problem are reported. In these researches, 200 kg of gadolinite were separated with regard to Pm by using one- and two-stage cation-exchange separations and nitrilotri-aceticacid for elution and Cu2+ as barrier-ion. The experiments yielded a beta-activity in the mixture-fractions between Sm and Nd. Because of its chemical behaviour and its half-thickness of 4.4 mg/cm2 that corresponds with a β-maximum energy of 230 keV (147Pm: 223 keV), this beta activity could be identified as 147Pm. For the lanthanide oxides of gadolinite a 147Pm content of 7.3 · 10 - 17% could be deduced, which agrees well with the value of 4.5 · 10 - 17% found by Erämetsä in 20 t lanthanide oxides of 6000 t apatite. The possible origin o f the 147Pm-isotope we found in gadolinite is discussed (fallout, actinium anomaly, spontaneous fission of 238U, and others), and the reaction 11,Id 2,6a 14 6Nd (n, y) is supposed to be the cause of the relatively short-lived isotope. The difference between the neutron flux (36 n/cm2 · min) that was deduced from the saturation activity of 147Pm, and the values of the neutron flux at the earth's surface or crust (0.001-0.24 n/cm2 · min) that up to now have been looked upon as probable by other authors, is also mentioned. Moreover, the question of the possible existence of a long-lived isomer of one of the known Pm-isotopes is rised.